1. Field of the Invention
The present invention relates to an image processing apparatus and an image processing method.
2. Description of the Related Art
An electrophotographic method is known as an image recording method used in a color image forming apparatus such as a color printer and a color copying machine. In the electrophotographic method, a latent image is formed on a photosensitive drum using a laser beam to develop the latent image by a charged color material (hereinafter referred to as “toner”). An image recording is performed such that the developed toner image is transferred to a transfer sheet to fix the image thereon.
Recently, for increasing a speed in forming an image in a color image forming apparatus using the electrophotographic method, tandem type color image forming apparatuses have increased that include the number of developing machines and the number of photosensitive drums (i.e., image recording units) corresponding to the number of toner colors and that sequentially transfers images of different colors on an image conveyance belt or a recording medium. In the tandem type color image forming apparatus, a plurality of factors for causing misregistration is known, and thus various methods are discussed for solving each of the factors.
Examples of the factors include unevenness and a mounting position deviation of a lens in a deflection scanning device and an assembling position deviation of the deflection scanning device to a color image forming apparatus body. Due to the positional deviation, an inclination or a bending of a scanning line occurs, and a degree of the bending (hereinafter referred to as “profile”) differs in each color for a color component of a toner, which causes the misregistration. Characteristics of the profile differs between image forming apparatus, i.e., between recording engines or between image recording units of different colors.
To solve an issue of the misregistration, for example, Japanese Patent Application Laid-Open No. 2004-170755 discusses a method in which degrees of an inclination and a bending of a scanning line are measured by an optical sensor and bitmap image data is corrected so as to offset the inclination and bending, thereby forming an image of the corrected image data. In this method, a mechanically adjustable member and an adjustment step upon assembling the apparatus are no longer required since the image data is processed to be electrically corrected. Therefore, downsizing of the color image forming apparatus can be achieved and the issue of the misregistration can be solved inexpensively.
The electrical misregistration correction includes a correction in one pixel unit and a correction in less than one pixel. In the correction in one pixel unit, pixels are offset by the one pixel in a sub-scanning direction according to correction amounts of the inclination and the bending. In a case where the above described method is used, the bending or the inclination caused due to the above described misregistration is about a range between 100 and 500 μm. In the image forming apparatus having a resolution of 600 dpi, an image memory for storing several tens of lines is required for the above described correction. In the below description, a position on the scanning line at which the pixel is offset is referred to as a changing point.
The correction in less than one pixel is performed such that a gradation value of the image data is adjusted by pixels before and the after a target pixel in the sub-scanning direction. The correction in less than one pixel can eliminate an unnatural step at a boundary of a changing point generated as a result of the correction in one pixel unit to smooth the image.
In a case where the above described smoothing processing is performed on an image that has been subjected to screen processing and is immediately before the printing, the smoothing processing is performed such that a pulse width modulation (PWM) is performed on a laser beam and a laser exposure time is gradually switched in the sub-scanning direction for smoothing the image. For example, in a case of a correction of 0.5 pixel, i.e., a correction in less than one pixel, the smoothing processing is realized by interpolation processing in which a half exposure is performed twice upwardly and downwardly in the sub-scanning direction.
Such interpolation processing can be performed only when a linear relationship is established between an exposure time of the PWM and an image density. Actually, a density obtained by the one time exposure of one pixel cannot be obtained in the two times exposure of 0.5 pixel in many cases. Therefore, if the density reproduced by the PWM cannot maintain the linearity to a density signal of a target to be processed, there exist two types of image data, i.e., image data that is preferably subjected to the above described interpolation processing and image data of which image quality may be degraded when it is corrected.
For example, with respect to patterns drawn by repeating the same designs or patterns (hereinafter referred to as “patterned image”), characters, and drawings which can be drawn by, for example, office word-processing software, the interpolation processing provided thereto, i.e., the smoothing processing thereof, can improve visibility of information. To the contrary, if the interpolation processing is performed at a changing point of a continuous tone image subjected to screen processing, an issue arises that density unevenness occurs only on the changing point due to correction processing, resulting in image quality deterioration. This is because, in a case where, for example, a line growth screen is used, the density appears to change in a macro perspective view since a line thickness in the screen changes on the changing point according to the interpolation processing. Further, in a case where an add-on image such as a copy-forgery-inhibited pattern is subjected to the interpolation processing, an effect of the interpolation processing might be lost. Therefore, the interpolation processing is not suitable to be performed for the add-on image.
Thus, if the interpolation processing using the PWM is performed, whether the interpolation processing is applied is determined according to an attribute of target image data. To resolve the above issue, such a method can be proposed that a continuous tone image determination unit and a patterned image determination unit are used to finally obtain an interpolation determination result from the determination results of these two units. In the continuous tone image determination unit, an image that is not to be interpolated can be determined. In the patterned image determination unit, an image that is to be interpolated can be determined.
For example, Japanese Patent Application Laid-Open No. 2003-274143 discusses a misregistration correction according to a geometric transformation with respect to an image after the screen processing. By inserting or deleting a pixel at a cycle at which no interference occurs with respect to a halftone dot cycle of a screen, the geometric transformation of the image is performed without causing unevenness and a moire of gradations. Such a minute transformation is realized by inserting or removing a pixel itself of a high resolution image without performing a pulse width modulation such as the PWM to partially shift the image in the main scanning direction or the sub-scanning direction.
As described above, it is difficult to perform the interpolation processing on any image while maintaining a good image quality of the image reproduced by a pulse width of the PWM in a state in which the linearity to a target density signal is hard to be established. Therefore, determination processing is required in providing the interpolation processing. However, in a case where an arbitrary image, e.g., a print image, is input from a user or an application, an erroneous determination may be made because of the determination processing.
In order to provide a high speed real time determination so as to catch up with a printing speed with respect to an arbitrary image, image determination processing needs to be performed by means of hardware since a real time determination cannot be performed at a satisfied speed by means of software. However, in attempting to perform the image determination processing by means of hardware, a circuit may require a complicated configuration depending on processing to be performed, which causes an increase of a size of the circuit. To the contrary, in attempting to perform the image determination processing by means of practical hardware, complicated determination processing cannot be performed in many cases.
In a case of performing the determination based on attribute information to be output from the user or the application about characters and photographs generated upon image rendering, similar to the above, there is a risk of the erroneous determination. In view of the image quality, in a case where the interpolation processing is not provided to the continuous tone image subjected to the screen processing, as described above, a step of one pixel that occurs at the changing point will be accepted. Accordingly, the step may be visually recognized as deterioration of the image depending on a type of the image.
An absolute amount of the corrected step needs to be minimized to the extent that is less than a certain value that a person hardly visually notices it. Since the absolute amount of the step of one pixel differs according to resolutions of printers, the step of one pixel needs to be divided into several steps according to the resolutions to generate steps less than one pixel. In a case where the geometric transformation is performed by shifting the image using the above described insertion or removal of the pixel, the size of the pixel needs to be as small as possible to the extent that a person hardly visually notices. Thus, high resolution is required. If image data after a pixel is inserted thereinto or removed therefrom is only shifted vertically with respect to the sub-scanning direction or the main scanning direction, a screen pattern is partially destroyed even if the pixel is inserted or removed by a cycle that avoids the interference.
Conventionally, there has been an issue that an image defect such as density unevenness occurs because the screen pattern that generates the cycle of screen of the image data is destroyed by the correction of the step less than one pixel.